基于BP-GRU组合模型的高速铁路短期客流预测

Short-time Passenger Flow Prediction of High-speed Railway Based on BP-GRU Combined Model

高速铁路短期客流预测需要高质量的准确度,提出BP-GRU深度学习组合模型以提高预测准确性。针对BP算法易陷入极小值的问题,构建BP-GRU训练模式与预测模式,从而提高BP算法的搜索能力。以2020年4月1日至2022年2月25日上海站客流为基础,从时序客流数据特点出发,完成了客流数据的标准化与重构。通过精细化参数设置,借助BP-GRU的不同模式,对高速铁路短期客流进行预测。通过与LSTM、GRU、BP-LSTM预测模型对比,采用合适的评价指标,结果表明BP-GRU组合模型比单一神经网络预测效果更好。BP-GRU比BP-LSTM拥有更加简单的网络的结构,预测结果基本一致的情况下,完成本次预测所花费的时间更少。

High-speed railway requires high quality accuracy to predict the short-time passenger flow,and the BP-GRU deep learning combined model is proposed to improve the accuracy of the prediction.Aiming at the problem that the BP algorithm is easy to fall into the extreme value,the training mode and prediction mode of the BP-GRU are constructed,to improve the search ability of the BP algorithm.Based on the passenger flow of Shanghai Station from April 1,2020 to February 25,2022,starting from the characteristics of time series passenger flow data,the standardization and reconstruction of passenger flow data is completed.Through refined parameter setting,with the help of different modes of BP-GRU,the short-time passenger flow of high-speed railway is predicted.By compared with LSTM,GRU,BP-LSTM predicted models,choose the appropriate evaluation indicators,the results showed that the BP-GRU has better than the single neural network prediction.BP-GRU has a simpler network structure than BP-LSTM,and the predicted results are basically consistent,and it takes less time to complete this prediction.